DOI: 10.1126/science.289.5477.284, 284 (2000);289 Science

et al.Charles J. VörösmartyPopulation GrowthGlobal Water Resources: Vulnerability from Climate Change and

This copy is for your personal, non-commercial use only.

clicking here.colleagues, clients, or customers by , you can order high-quality copies for yourIf you wish to distribute this article to others

  here.following the guidelines

can be obtained byPermission to republish or repurpose articles or portions of articles

  ): June 7, 2013 www.sciencemag.org (this information is current as of

The following resources related to this article are available online at

http://www.sciencemag.org/content/289/5477/284.full.htmlversion of this article at:

including high-resolution figures, can be found in the onlineUpdated information and services,

http://www.sciencemag.org/content/289/5477/284.full.html#ref-list-1, 1 of which can be accessed free:cites 9 articlesThis article

296 article(s) on the ISI Web of Sciencecited by This article has been

http://www.sciencemag.org/content/289/5477/284.full.html#related-urls26 articles hosted by HighWire Press; see:cited by This article has been

registered trademark of AAAS. is aScience2000 by the American Association for the Advancement of Science; all rights reserved. The title

CopyrightAmerican Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005. (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by theScience

on

June

7, 2

013

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

Global Water Resources:Vulnerability from Climate

Change and Population GrowthCharles J. Vorosmarty,1,2,4,5* Pamela Green,1,2,4

Joseph Salisbury,1,3,4 Richard B. Lammers1,2,4

The future adequacy of freshwater resources is difficult to assess, owing to acomplex and rapidly changing geography of water supply and use. Numericalexperiments combining climate model outputs, water budgets, and socioeco-nomic information along digitized river networks demonstrate that (i) a largeproportion of the world’s population is currently experiencing water stress and(ii) rising water demands greatly outweigh greenhouse warming in defining thestate of global water systems to 2025. Consideration of direct human impactson global water supply remains a poorly articulated but potentially importantfacet of the larger global change question.

Greenhouse warming continues to dominate theworld’s science and policy agenda on globalchange. One fundamental concern is the impactof this climate change on water supply (1, 2).The question of how human society directlyinfluences the state of the terrestrial water cyclehas received much less attention, despite thepresence of the socioeconomic equivalent ofthe Mauna Loa curve, namely, rapid populationgrowth and economic development. Our goal inthis report is to identify the contributions ofclimate change, human development, and theircombination to the future state of global waterresources.

Assessments of water vulnerability tradi-tionally have been cast at the country or region-al scale (2–5). Although recent work has fo-cused on individual drainage basins and subba-sins (1, 6, 7), to the best of our knowledge, noglobal-scale study has articulated the geograph-ic linkage of water supply to water demanddefined by runoff and its passage through rivernetworks. We present a high-resolution geogra-phy of water use and availability, analyzing thevulnerability of water resource infrastructure(8) to future climate change, population growthand migration, and industrial development be-tween 1985 and 2025. We consider explicitlyhow the topology of river systems determinesthe character of sustainable water supply and itsuse by humans.

Mean annual surface and subsurface (shal-low aquifer) runoff, accumulated as river dis-charge (Q), is assumed to constitute the sustain-able water supply to which local human popu-lations have access (9). We mapped the distri-bution of population with respect to relative

water demand (RWD) defined as the ratio ofwater withdrawal or water use to discharge. Weconsider the domestic and industrial sectors(DI/Q), irrigated agriculture (A/Q), and theircombination (DIA/Q) on a mean annual basis.Each ratio determines the degree to which hu-mans interact with sustainable water supply andprovides a local index of water stress. Values onthe order of 0.2 to 0.4 indicate medium to highstress, whereas those greater than 0.4 reflectconditions of severe water limitation (10). Wealso constructed a water reuse index (SDIA/Q),defined as the ratio of aggregate upstream wateruse relative to discharge. We consider vulnera-bility with respect to sustainable water resourcesonly. We make no explicit tabulation of nonsus-tainable supplies or withdrawals, such as themining of groundwater, although we can drawinferences about such activities by analyzingRWD. We do not explicitly model human ad-aptation to climate change or development pres-sure, but we do incorporate estimates of futurewater use efficiency offered in other studies.

A recent version of the Water Balance Mod-el (WBM) (11) was used to compute contem-porary and future runoff at 309 grid resolution(latitude by longitude). Runoff fields were con-strained by monitoring data, and converted todischarge by integrating along digitized rivers(12, 13). Climate change fields were from theCanadian Climate Center general circulationmodel CGCM1 and Hadley Center circulationmodel HadCM2 used in the current Intergovern-mental Panel on Climate Change (IPCC) assess-ment (14). Global means for contemporary(1961–90) runoff and river discharge were com-puted by the WBM using off-line atmosphericforcings from HadCM2 and CGCM1. Predic-tions were in substantial agreement with runofffields based on observed discharge (13, 15).Results from HadCM2/WBM and CGCM1/WBM were used to predict incremental differ-ences between contemporary and future runoffand discharge for individual grid cells. These

differences were then applied to a baseline (13)to generate the future patterns of runoff (16).Mean global runoff varied in response to climatechange from an increase of ,1 mm year21

(HadCM2/WBM) to a decrease of 17 mmyear21 (CGCM1/WBM) (17). With each runofffield, more substantial changes could be foundat local and regional scales. CGCM1/WBMgave the strongest climate change signal, and weuse it to exemplify key findings derived fromboth models.

Domestic and industrial water demand wasdetermined by population and per capita usestatistics. The geography of contemporary ur-ban and rural population was developed from a1-km data set (18). Future population distribu-tion was determined from projections of thepercent change in total, rural, and urban popu-lation from 1985 to 2025 (19) applied to the1-km urban and rural population maps. Coun-try-level water withdrawal statistics (19) wereused to estimate contemporary water demands,but they first required standardization and spa-tial disaggregation (20). The geography of ag-ricultural water demand was computed fromirrigated land area and national use statistics(21). Future demands for all sectors were basedon population growth, economic development,and projected changes in water use efficiency(22). Water withdrawals at 309 resolution weregeographically linked to digital river networksand corresponding discharge estimates.

The contemporary condition is represent-ed by 1985, the year that is most compatiblewith the time span represented by the runoffclimatology and historical water use statis-tics. Against this benchmark we formulatedthree scenarios to quantify the contributionsof climate change and development pressureto the degree of relative water demand in2025. The first scenario (Sc1) varied climatebut fixed the magnitude and spatial distribu-tion of human population and water with-drawals at 1985 levels. Sc2 applied projectedwater demands for 2025 but used runoff anddischarge based on contemporary climate.Sc3 changed both climate and water demand.Total water use per capita is projected todecrease from 640 to 580 m3 year21 between1985 and 2025. The impacts of human devel-opment under Sc2 and Sc3 will thereforegenerally reflect population growth and mi-gration as opposed to intensification of wateruse, though results will be location specific.In relation to (5), our calculation of globalwater use in 2025 is conservative, 4700 km3

year21 compared to 5200 km3 year21.We compared our calculations to country-

level data typically applied in global water as-sessments. Our national-scale aggregates ofgridded DIA/Q and a recent global assessmentby the United Nations (10) place almost thesame fraction of the world’s 1995 populationunder similar levels of water stress (Table 1). Inboth studies, one-third of the total population of

1Water Systems Analysis Group, 2Complex SystemsResearch Center, 3Ocean Processes Analytical Labo-ratory, 4Institute for the Study of Earth, Oceans, andSpace, 5Earth Sciences Department, University ofNew Hampshire, Durham, NH 03824, USA.

*To whom correspondence should be addressed.

R E P O R T S

14 JULY 2000 VOL 289 SCIENCE www.sciencemag.org284

on

June

7, 2

013

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

5.7 billion lives under conditions of relativewater scarcity (DIA/Q . 0.2), and ;450 mil-lion people are under severe water stress (DIA/Q . 0.4). A summary based on individual gridcells (Table 1) shows that a much larger pop-ulation (an additional 1.3 billion) now livesunder a high degree of water stress that nation-al-level totals, especially for large countries, failto articulate. Use of 309 grids (n 5 59,132)captures much more of the spatial heterogeneityin water use, discharge, and RWD (Fig. 1).Water stress transcends national boundaries andis apparent today across arid and semiarid re-gions as well as in many densely populatedparts of the humid tropics and temperate zone.

We find that the primary determinants ofchanging levels of RWD, and hence vulner-ability to water stress, through the early part

of this century will be the growth and eco-nomic development of human population.We base this conclusion on contrasts betweenthe cumulative distributions of global popu-lation, ranked by DI/Q and A/Q, for each ofthe scenarios tested (Table 2). UnderCGCM1/WBM, we see almost no differencebetween cumulative population distributionsrepresented by the contemporary baseline andby climate change scenario Sc1. In contrast,Sc2 shows a large effect from human devel-opment with substantial increases over 1985in accumulated population for all levels ofDI/Q. The additional climatic effects repre-sented by Sc3 fail to elicit a substantial de-parture from the Sc2 distribution. Althoughmore people are predicted in 2025 to beliving in relatively water-rich areas, under

Sc3 the highly vulnerable population withDI/Q . 0.4 increases to .2 billion, an 85%increase in relation to the vulnerable popula-tion in 1985. This condition is determinedalmost exclusively by population and devel-opment pressure.

For agriculture, overall results are similar(Table 3). The population distribution here re-fers to the number of people dependent onirrigated water withdrawals (21), and changesin either remote demand or local available dis-charge influence A/Q under contrasting scenar-ios. The effect of Sc1 produces little changefrom 1985, and the aggregate impact of increas-ing water demands under Sc2 and Sc3 is appar-ent. For 1985, we estimate that almost 2 billion

Table 1. Contemporary world population living under progressive levels of relative total water demand(DIA/Q). The thresholds and definitions of water stress are as defined by the United Nations (UN) (10).Results shown here contrast national-level summaries (;102 entries) against grid-based tabulations(;104 to 105 entries). Indices given are for 1995.

Water stressDIA/Q

(unitless)

Total population (billions)

Country-level Grid-based

UN This study* This study

Low ,0.1 1.72 1.95 3.16Moderate 0.1 to 0.2 2.08 1.73 0.38Medium-high 0.2 to 0.4 1.44 1.54 0.37High .0.4 0.46 0.45 1.76

*Total water demand, runoff, and population at 309 grid spatial resolution were each summed to the national scale, andcorresponding aggregates were then computed.

Fig. 1. The global distribution of population in 1985 with respect to the relativewater stress threshold of DIA/Q 5 0.4 indicating severe water scarcity (10). A309 spatial resolution is used. This mapping reflects a mean global runoff of;40,000 km3 year21 and aggregate water withdrawals of 3100 km3 year21.These estimates are highly dependent on contemporary water use statistics,

which reflect a degree of uncertainty. Recent reviews (5, 36) show year 2000global water withdrawals from assessments made even as late as 1987 to varyby .1300 km3 year21. National-level water use statistics (18) for some coun-tries are decades old. Runoff estimates for some regions may also be biased(9,13). Results should be viewed with appropriate caution.

Table 2. Cumulative distribution of worldwidepopulation with respect to ranked values of rela-tive water demand for domestic and industrialsectors (DI/Q) generated by the CGCM1/WBMmodel. Each entry represents the population at orexceeding the indicated DIA/Q level from 309 res-olution data.

DI/Q(unitless)

Cumulative population (billions)

Con-temporary

Sc1 Sc2 Sc3

1.0 0.9 0.8 1.9 1.60.4 1.2 1.0 2.4 2.20.2 1.4 1.3 2.8 2.70.1 1.7 1.6 3.2 3.20.01 2.9 2.9 5.4 5.40.001 4.1 4.1 7.0 7.00.000 4.8 4.8 8.0 8.0

R E P O R T S

www.sciencemag.org SCIENCE VOL 289 14 JULY 2000 285

on

June

7, 2

013

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

people are dependent on irrigated lands withA/Q values of .0.4. For 2025, under Sc3 thisnumber rises to .3 billion. Irrigation thus sup-ports 40% of the population in 1985 (and willpotentially support the same percentage in2025) on cropland with A/Q values of .0.4,suggesting a substantial nonsustainable wateruse and major global vulnerability, even underpresent-day conditions.

Our findings are further supported by cal-culations expressed as continental- and glo-bal-scale totals (Table 4). For the globe, cli-mate change under Sc1 increased DIA/Q val-ues by ,5%. In contrast, rising water de-mands alone (Sc2) increased DIA/Q by 50%,whereas Sc3 combining both climate and de-velopment effects produced relative increasesof 60%. Over individual continents, climate-induced changes in DIA/Q varied from a 4%decrease to a 12% increase, which were in allcases much smaller than changes correspond-ing to population and economic growth.

Continental- and global-scale summariesmask potentially important regional patterns ofwater abundance and scarcity. We accumulatedwater demand and water supply and calculatedSDIA/Q along main-stem rivers to establish an

aggregate imprint of water use intensity andcompetition across watersheds. Even rivers inclose proximity show distinct patterns ofSDIA/Q and of sensitivity to future changes inclimate and water demand (Fig. 2). The ChangJiang River (China) follows a pattern of sensi-tivity under which both climate change andpopulation pressures increase the water reuseindex along virtually the entire main stem. Un-der Sc3, we see a severalfold increase inSDIA/Q over contemporary conditions, an im-pact determined in large measure by climatechange. The neighboring Yellow River alsodisplays a progressive intensification ofSDIA/Q in the downstream direction but withan aggregate use of water well in excess of theentire basin’s discharge, even for the baselinecondition. Future development pressure (Sc2)exacerbates the situation, whereas climatechange has an apparent beneficial effect by

lowering values of SDIA/Q over the entiremain stem, thereby counteracting the increasesassociated with future population growth. As aresult, future SDIA/Q values are lowered sub-stantially. Despite the projected improvement,there is likely to be a sustained and severepressure on water supplies in this basin. Con-temporary conditions along this river are al-ready more severe than indicated, because ofrapid increases in water use and decreases indischarge into the 1990s (23), which are notreflected by the 1985 benchmark.

When results are examined at the grid-celllevel, an even richer set of responses emerges.The increase or decrease in SDIA/Q for eachscenario in relation to conditions in 1985 wasused to develop a geography of changing rela-tive water demand (Fig. 3). Climate changealone (Sc1) produces a mixture of responses,both positive and negative, that is highly region

Fig. 2. The imprint ofaccumulated relativewater demand from allsectors (SDIA/Q) plot-ted as a function ofdownstream distancealong two major riversin eastern Asia. The con-temporary setting iscontrasted against thethree scenarios of po-tential conditions in2025 simulated byCGCM1/WBM. Trajec-tories are unique for in-dividual main-stem riv-ers and involve a com-plex interplay betweenthe geography of riverdischarge and wateruse. An increase in thisindex along the down-stream direction ac-companies an increasein accumulated waterdemand, a decrease indischarge, or both,whereas a lowering ofthe curve reflects dilu-tion from local runoff or less impacted tributaries. SDIA/Q is an index of water competition andreuse as well as a surrogate for potential water quality problems.

Table 3. Same as Table 2, except showing thecumulative distribution of worldwide populationthat is dependent on contemporary water usefor irrigated agriculture at different levels ofrelative demand (A/Q). Dependent populationswere linked to irrigation water demands withinindividual countries; table entries are derivedfrom gridded 309 data.

A/Q(unitless)

Cumulative dependentpopulation (billions)

Con-temporary

Sc1 Sc2 Sc3

1.0 1.5 1.4 2.7 2.70.4 1.9 1.8 3.4 3.30.2 2.2 2.1 4.0 3.90.1 2.6 2.5 4.7 4.70.01 3.8 3.8 6.7 6.60.001 4.5 4.5 7.6 7.60.000 4.8 4.8 8.0 8.0

Table 4. Continental and global summaries for population, irrigable land,sustainable water supply defined as discharge (Q), and relative all-sectorwater demand (DIA/Q) tabulated for the contemporary condition and

simulated by CGCM1/WBM. Percentages assigned to the change in DIA/Q(DDIA/Q) are relative to the 1985 contemporary baseline.

Area

Population(millions) Irrigated

cropland(1000 km2)

ObservedQ

(km3 year21)

ContemporaryDIA/Q

(unitless)

2025Q

(km3 year21)

Predicted DDIA/Q (%)

1985 2025 Sc1 Sc2 Sc3

Africa 543 1440 118 4,520 0.032 4,100 10 73 92Asia 2930 4800 1690 13,700 0.129 13,300 2.3 60 66Australia/Oceania 22 33 26 714 0.025 692 2.0 30 44Europe 667 682 273 2,770 0.154 2,790 21.9 30 31North America 395 601 317 5,890 0.105 5,870 24.4 23 28South America 267 454 95 11,700 0.009 10,400 12 93 121

Globe 4830 8010 2520 39,300 0.078 37,100 4.1 50 61

R E P O R T S

14 JULY 2000 VOL 289 SCIENCE www.sciencemag.org286

on

June

7, 2

013

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

specific. Expanded water use by itself (Sc2)increases relative SDIA/Q for broad regions ofthe globe, although small clusters of grid cellsshowing relative decreases appear in areas ofrural-to-urban migration, as in Russia. Thelarge continental areas with elevated SDIA/Qvalues under Sc3 reflect well the patterns ofincrease associated with Sc2. Interactions be-tween population growth and climate changeresult in some notable net decreases in SDIA/Q, which are large enough to reverse the rela-tive water scarcity suggested by Sc2, as inMexico and much of central Asia. The overall

pattern, however, is one of pandemic increase.The major increases in relative water de-

mand documented here reveal that much ofthe world will face substantial challenges towater infrastructure and associated water ser-vices. Potentially large economic costs arelikely to be associated with the implementa-tion of response strategies (e.g., expansion offacilities, new water-pricing policies, innova-tive technology, and mismanagement) or theconsequences of inaction (e.g., deteriorationof water quality and reduction in irrigatedcrop yields) (24, 25). Where sustainable wa-

ter supplies are at a premium, the challengesalso include curtailment of economic activi-ties, abandonment of existing water facilities,mass migration, and conflict in internationalriver basins (25–27).

Many parts of the developing world willexperience large increases in relative waterdemand. In water-rich areas such as the wettropics, the challenge will not be in providingadequate quantities of water, but in providingclean supplies that minimize public healthproblems (28). Arid and semiarid regionsface the additional challenge of absolute wa-ter scarcity. Projected increases in scarcitywill be focused on rapidly expanding cities.Much of the world’s population growth overthe next few decades will occur in urbanareas, which are projected to double in size tonear 5 billion between 1995 and 2025 (29)and face major challenges in coping withincreased water pollution and incidence ofwaterborne disease (5, 10, 19, 25, 29).

We conclude that impending global-scalechanges in population and economic develop-ment over the next 25 years will dictate thefuture relation between water supply and de-mand to a much greater degree than will chang-es in mean climate. To secure a more completepicture of future water vulnerabilities, it will benecessary to consider interactions among cli-mate change and variability, land surface andgroundwater hydrology, water engineering, andhuman systems, including societal adaptationsto water scarcity [see (30, 31)]. Pursuit of thisquestion will be limited by outdated and non-existent socioeconomic data and informationfrom a progressively deteriorating global net-work of hydrometric monitoring stations (32)unless a vigorous commitment is made by thewater sciences community to collect, standard-ize, and widely disseminate such information.In light of our findings, an integrated approachbringing together the climate change, water re-sources, and socioeconomic communities ap-pears essential to future progress.

References and Notes1. N. Arnell et al., in Climate Change 1995: Impacts,

Adaptations, and Mitigation of Climate Change, R. T.Watson et al., Eds. (Cambridge Univ. Press, Cam-bridge, 1996), pp. 325–363.

2. Z. Kaczmarek et al., in Climate Change 1995: Impacts,Adaptations, and Mitigation of Climate Change, R. T.Watson et al., Eds. (Cambridge Univ. Press, Cam-bridge, 1996), pp. 469–486.

3. M. Falkenmark, Water Int. 16, 229 (1991).4. M. I. L’vovich and G. F. White, in The Earth as Trans-

formed by Human Action, B. L. Turner et al., Eds.(Cambridge Univ. Press, Cambridge, 1990), pp. 235–252.

5. I. Shiklomanov, Ed., Assessment of Water Resourcesand Water Availability in the World: Scientific andTechnical Report (State Hydrological Institute, St.Petersburg, Russia, 1996).

6. J. Alcamo et al., in World Water Scenarios: Analyses,F. R. Rijsberman, Ed. (Earthscan, London, 2000), pp.204–242.

7. J. C. van Dam, Ed., Impacts of Climate Change andClimate Variability on Hydrological Regimes (Cam-bridge Univ. Press, Cambridge, 1999).

Fig. 3. Maps of the change in water reuse index (SDIA/Q) predicted by the CGCM1/WBM modelconfiguration under Sc1 (climate change alone), Sc2 (population and economic development only),and Sc3 (both effects). Changes in the ratio of scenario-specific SDIA/Q (SDIA/QScenario) relativeto contemporary (SDIA/QBase) conditions are shown. A threshold of 620% is used to highlightareas of substantial change.

R E P O R T S

www.sciencemag.org SCIENCE VOL 289 14 JULY 2000 287

on

June

7, 2

013

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

8. “Water resource infrastructure” refers to watersource, distribution, and treatment systems. We as-sume that wherever there is a resident human pop-ulation or irrigated cropland, there will be a corre-sponding water infrastructure. Changes in water de-mand due to population growth and industrializationor in water supply due to climate change will definethe vulnerability of water infrastructure and the hu-man population that is dependent on these systems.

9. S. L. Postel et al., Science 271, 785 (1996).10. United Nations, Comprehensive Assessment of the

Freshwater Resources of the World (overview docu-ment) (World Meteorological Organization, Geneva,1997).

11. C. J. Vorosmarty, C. A. Federer, A. Schloss, J. Hydrol.207, 147 (1998).

12. C. J. Vorosmarty, B. Fekete, M. Meybeck, R. Lammers,Global Biogeochem. Cycles 14, 599 (2000).

13. B. M. Fekete et al., Global, Composite Runoff FieldsBased on Observed River Discharge and SimulatedWater Balances, Report 22 (World MeteorologicalOrganization–Global Runoff Data Center, Koblenz,Germany, 1999).

14. Data are from the IPCC Data Distribution Centre, Deut-sches Klimarechenzentrum (Max-Planck-Institut) inHamburg, Germany, and the Climatic Research Unit atthe University of East Anglia in Norwich, UK.CGCM1GSa1 and HadCM2GSa1 (GSa, ensemble ofgreenhouse gas plus sulfate aerosol integrations) sce-narios were obtained from http://ipcc-ddc.cru.uea.ac.uk/cru_data/datadownload/download_index.html.Scenarios represent a 1% per year increase in CO2-equivalent forcing and sulfate aerosol dampening. Orig-inal data at 3.75° by 3.75° (latitude by longitude) forCGCM1 and at 2.5° by 3.75° for HadCM2 were bilin-early interpolated to 309 resolution. Monthly forcingswere applied to the WBM, and a statistically equivalentdaily time step was used to integrate over time andcompute water budget variables, including runoff.

15. Simulated water budgets combined with dischargedata from several hundred recording stations in (13)yielded a mean global runoff of 300 mm year21 or adischarge of 39,300 km3 year21; CGCM1/WBM com-puted respective values of 319 mm year21 and41,900 km3 year21, whereas HadCM2/WBM gave302 mm year21 and 39,600 km3 year21, respectively.

16. The approach taken is that used in climate impactstudies on net primary production by VEMAP Mem-bers [Global Biogeochem. Cycles 9, 407 (1995)].

17. The values are statistically significant (P , 1 3 1026)with the Wilcoxon sign test.

18. A 1-km gridded polygon file [Arc World Supplement, 1:3M scale digital map (ESRI, Redlands, CA, 1995)] definedthe spatial extent of 242 countries for which country-level population statistics were available (19). We de-fined urban spatial extents as a set of geographicallyreferenced city polygons with demographic data (n 51858) (33) and distributed the remaining country-levelurban population evenly across 1-km pixels classified ascity lights from remote sensing (34). Lacking digital datato the contrary, we distributed rural population uni-formly among digitized points representing populatedplaces [Digital Chart of the World, 1:1 M scale digitalmap (ESRI, Redlands, CA, 1993)] falling outside of urbanspatial extents. A total of 155 countries simultaneouslyshowed water demand data and discharges greater thanzero and fell within our 309 digitized land mass. Theremaining 87 countries were mostly small islands andwere not considered. For the contemporary setting, weaccount for 99.7% of the global population (19); 98.4%of the total is assigned water use statistics.

19. World Resources: A Guide to the Global Environment1998 –99 (World Resources Institute, Washington,DC, 1998).

20. National and sectoral water use statistics were from(19). The mean reporting year was 1986, but the rangewas from 1970 to 1995. National statistics were nor-malized to year 1985 by applying usage trends recordedin corresponding regional time series (5). Domestic wa-ter demand was computed on a per capita basis foreach country and distributed geographically with re-spect to the 1-km total population field. Industrialusage was applied in proportion to urban population.Grid-based aggregates at 309 resolution were then de-termined for domestic plus industrial water demand.

21. Country-level totals for agricultural water demandwere distributed onto 309 grid cells on the basis ofthe fraction of each grid cell classified as irrigatedland from (35) and prorated on the basis of the ratioof unrealized potential evapotranspiration (i.e., thepotential minus the estimated actual) to the poten-tial from (13). Irrigation-dependent population wasdetermined by proportionally assigning national-levelpopulation to the corresponding irrigated areas ineach country. We reason that entire national popu-lations (and not simply local farmers and agribusi-ness) benefit from the food and fiber (destined fordomestic or export markets) and income producedfrom irrigated land. A/Q uses mean annual discharge.These relative water demand estimates are thus con-servative and assume highly effective storage of sur-face water for irrigation, such as through reservoirimpoundment. We consider irrigated agriculture be-cause it is a major component of water resourceinfrastructure that is subject to changes in the avail-ability of net runoff. Rain-fed agriculture falls outsidethis definition, and we have not treated it here.

22. Rates of increase in water demand to 2025 fromregional estimates (5) were applied to the 1985water withdrawal data set. Future changes in pop-ulation and urban-to-rural ratios (19) were used toshift the geography of water demands. The distri-bution of irrigable lands was fixed to that observedunder contemporary conditions. Projected waterwithdrawals in (5) are dependent on water useefficiencies that both increase and decrease fordifferent parts of the world. These estimates weremade through extensive consultation of country-level studies and trend analysis based on per unitagricultural, municipal, and industrial water with-drawals; assumptions regarding future technologyadoption; and economic capacity to institute effi-ciency changes.

23. J. Milliman and R. Mei-e, in Climate Change: Impacton Coastal Habitation, D. Eisma, Ed. (CRC Press, BocaRaton, FL, 1995), pp. 57–83.

24. S. Postel, Interciencia 10, 290 (1985).25. P. Gleick, The World’s Water: The Biennial Report on

Freshwater Resources (1998 –99) (Island, Washing-ton, DC, 1998).

26. T. Homer-Dixon, Int. Secur. 19, 5 (1994).27. M. Falkenmark and J. Rockstrom, Ambio 22, 427

(1993).28. M. Bonell et al., Hydrology and Water Management in

the Humid Tropics (Cambridge Univ. Press, Cam-bridge, 1993).

29. World Resources: A Guide to the Global Environment1996 –97 (World Resources Institute, Washington,DC, 1996).

30. D. Conway et al., Ambio 25, 336 (1996).31. K. M. Strzepek et al, in World Water Scenarios: Anal-

yses, F. R. Rijsberman, Ed. (Earthscan, London, 2000),pp. 120–159.

32. J. C. Rodda, in Water: A Looming Crisis (InternationalHydrological Program, United Nations Educational,Scientific, and Cultural Organization, Paris, 1998).

33. W. Tobler et al., The Global Demography Project,Technical Report TR-95-6 (National Center for Geo-graphic Information and Analysis, Santa Barbara, CA,1995).

34. C. Elvidge et al., Int. J. Remote Sens. 18, 1373 (1997).35. P. Doll and S. Siebert, A Digital Global Map of Irrigat-

ed Areas, Report A9901 (University of Kassel, Kassel,Germany, 1999).

36. P. Gleick, in World Water Scenarios: Analyses, F. R.Rijsberman, Ed. (Earthscan, London, 2000), pp. 27–37.

37. Support for this work was through the Institute forthe Study of Earth, Oceans, and Space (University ofNew Hampshire); NASA Earth Observing System(grant NAG5-6137); NSF Division of AtmosphericSciences (grant ATM-9707953); Office of Polar Pro-grams (grant OPP-9524740); NASA Tropical RainfallMonitoring Mission (grant NAG5-4785); and the U.S.Department of Energy (DE-FG02-92ER61473). Weacknowledge the efforts of B. Fekete and S. Glidden inhelping to develop some of the geographically refer-enced databases used in this study. We also thankthree anonymous reviewers for their comments.

2 February 2000; accepted 3 May 2000

Overpressure and Fluid Flow inthe New Jersey ContinentalSlope: Implications for Slope

Failure and Cold SeepsBrandon Dugan* and Peter B. Flemings

Miocene through Pleistocene sediments on the New Jersey continental slope(Ocean Drilling Program Site 1073) are undercompacted (porosity between 40and 65%) to 640 meters below the sea floor, and this is interpreted to recordfluid pressures that reach 95% of the lithostatic stress. A two-dimensionalmodel, where rapid Pleistocene sedimentation loads permeable sandy silt ofMiocene age, successfully predicts the observed pressures. The model describeshow lateral pressure equilibration in permeable beds produces fluid pressuresthat approach the lithostatic stress where overburden is thin. This transfer ofpressure may cause slope failure and drive cold seeps on passive margins aroundthe world.

Rapid sediment loading (.1 mm year21) isdocumented as a source of overpressure (P*,pressure in excess of hydrostatic) in basins

around the world (1, 2). A suite of modelsdescribe how overpressure is generated duringrapid deposition (3–6). These models quantifythe rock properties and sedimentation rates re-quired to generate and maintain overpressure.Mass and volume measurements of wet and drycore samples provide porosity data (7) that weuse to document overpressures on the New

503 Deike Building, Department of Geosciences, PennState University, University Park, PA 16802, USA.

*To whom correspondence should be addressed. E-mail: dugan@geosc.psu.edu

R E P O R T S

14 JULY 2000 VOL 289 SCIENCE www.sciencemag.org288

on

June

7, 2

013

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

DOI: 10.1126/science.1128845, 1068 (2006);313 Science

Taikan Oki and Shinjiro KanaeGlobal Hydrological Cycles and World Water Resources

This copy is for your personal, non-commercial use only.

clicking here.colleagues, clients, or customers by , you can order high-quality copies for yourIf you wish to distribute this article to others

  here.following the guidelines

can be obtained byPermission to republish or repurpose articles or portions of articles

  ): June 7, 2013 www.sciencemag.org (this information is current as of

The following resources related to this article are available online at

http://www.sciencemag.org/content/313/5790/1068.full.htmlversion of this article at:

including high-resolution figures, can be found in the onlineUpdated information and services,

http://www.sciencemag.org/content/313/5790/1068.full.html#relatedfound at:

can berelated to this article A list of selected additional articles on the Science Web sites

http://www.sciencemag.org/content/313/5790/1068.full.html#ref-list-1, 4 of which can be accessed free:cites 17 articlesThis article

87 article(s) on the ISI Web of Sciencecited by This article has been

http://www.sciencemag.org/content/313/5790/1068.full.html#related-urls15 articles hosted by HighWire Press; see:cited by This article has been

http://www.sciencemag.org/cgi/collection/atmosAtmospheric Science

subject collections:This article appears in the following

registered trademark of AAAS. is aScience2006 by the American Association for the Advancement of Science; all rights reserved. The title

CopyrightAmerican Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005. (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by theScience

on

June

7, 2

013

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

REVIEW

Global Hydrological Cycles andWorld Water ResourcesTaikan Oki1,2,3*† and Shinjiro Kanae1,4*

Water is a naturally circulating resource that is constantly recharged. Therefore, even though thestocks of water in natural and artificial reservoirs are helpful to increase the available water resourcesfor human society, the flow of water should be the main focus in water resources assessments. Theclimate system puts an upper limit on the circulation rate of available renewable freshwater resources(RFWR). Although current global withdrawals are well below the upper limit, more than two billionpeople live in highly water-stressed areas because of the uneven distribution of RFWR in time andspace. Climate change is expected to accelerate water cycles and thereby increase the available RFWR.This would slow down the increase of people living under water stress; however, changes in seasonalpatterns and increasing probability of extreme events may offset this effect. Reducing currentvulnerability will be the first step to prepare for such anticipated changes.

All organisms, including humans, re-

quire water for their survival. There-

fore, ensuring that adequate supplies

of water are available is essential for human

well-being. Although our planet is often called

the BBlue Planet,[ warnings of increasing water

scarcity in the world are common. However,

unlike oil, water circulates, forming closed

hydrologic cycles. The amount of water will

not diminish on shorter than geological time

scales (1). Given this background, how could

water scarcity become a widespread reality

within a few decades (2)?

A common explanation is that even though

there is a lot of water on Earth, only about 2.5%

is fresh water, and because most of that water is

stored as glaciers or deep groundwater, only a

small amount of water is easily accessible. This

answer is only partly correct: Rather than looking

only at the stocks ofwater resources, assessments

should concentrate mainly on the flows (Fig. 1)

(1, 3–5). The amount of water stored in all the

rivers in the world is only 2000 km3, much

less than the annual water withdrawal of 3800

km3/year (Fig. 1). Clearly, a more adequate mea-

sure of water availability is the 45,500 km3/year

of annual discharge, which flows mainly through

the rivers from continents to the sea.

What Is the Meaning of a Circulating Resource?

Unlike most other natural resources, water cir-

culates naturally. When it evaporates, it changes

from liquid to gas and eventually recondenses as

a liquid. Water assimilated during photosyn-

thesis becomes part of carbohydrates stored in

plants, but ultimately reverts to water again by

decomposition.

When used, water loses properties such as

purity, heat content, and potential gravitational

energy, but eventually, most degraded water

resources are refreshed by natural processes in

the hydrological cycle, which is mostly driven

by solar energy. When considering water flux as

the most relevant measure of water resources,

the speed of water circulation becomes crucial.

Mean residence times of water molecules—i.e.,

how long they stay in a given reservoir—can be

estimated by dividing the volume of the reser-

voir by the mean flux into and out of it. For

rivers unaffected by human interventions, the

mean residential time of water is about two and a

half weeks (1). In contrast, the recharge rate of

some groundwater aquifers is very slow, and

the mean residential time is considered to be

hundreds or even thousands of years. When

water is extracted from such an aquifer, it will

take a very long time, measured on a human

time scale, to return to the original volume

stored; in practice, that water is exhausted once

it has been used. Because it took so long to

accumulate, the groundwater in such aquifers is

sometimes called fossil water.

How Much Renewable Fresh WaterIs Available?

Can human demand for water be fully met by

using only circulating renewable freshwater

resources (RFWR)? The answer is both yes

and no. Even though RFWR is naturally re-

cycled, the circulation rate is determined by the

climate system, and there is an upper limit to the

amount of RFWR available to human society.

On the global scale, current withdrawals are well

below this limit, and if the water cycle is man-

aged wisely, RFWR can cover human demand

far into the future. Appropriate water manage-

ment is a crucial point.

Conventional engineers of water resources

consider the water withdrawn from surface and

groundwater as water resources and evapo-

transpiration as a loss of water from the pre-

cipitatedwater. In that sense, precipitationminus

evapotranspiration over land is a measure of the

maximum available RFWR. The major part of

this available RFWR is surface water, particu-

larly river discharge. However, some part of the

water, approximately 10% of total river dis-

charge (6), infiltrates to deep underground and

will never appear as surface water but discharge

into the ocean directly from groundwater.

In contrast to the conventional view, it has

been noted that evapotranspiration from non-

irrigated cropland also is a water resource that is

beneficial to society (7). To distinguish between

this kind of resource and conventional re-

sources, evapotranspiration flow has been named

green water, and conventional withdrawal from

rivers and groundwater has been named blue

water (7).

About 3800 km3/year of RFWR (blue water)

is currently withdrawn by human beings, and

that accounts for less than 10% of the maxi-

mum available RFWR in the world (Fig. 1).

Evapotranspiration is estimated to be 7600

km3/year from cropland and 14,400 km3/year

from permanent grazing land. Cropland and

grazing land account for about one-third of the

total terrestrial evapotranspiration.

Can We Use All the RFWR?

Why should we be concerned about water

scarcity when presently only 10% of maximum

available blue water and 30% of green water

resources are used? The reason is the high

variability of water resource availability in time

and space (8). For example, the monthly mean

discharge at the Obidos station in the Amazon

River differs by a factor of 2 between the highest

and the lowest months, even for climatologically

averaged values. River discharge is more varia-

ble in smaller river basins in general, and daily

river discharge is, of course, more variable than

monthly river discharge. Because of this tempo-

ral variability, it is impractical to use 100% of

the available RFWR for human society. Flow

during floods and wet seasons cannot be used

during the low flow seasons unless storage sys-

tems are in place. That is why there are millions

of artificial reservoirs, lakes, and ponds in the

world and why most of the major rivers are

regulated (9). Total capacity of this artificial

storage is estimated to 7200 km3 (10), about

twice the annual water withdrawal.

Another reason that RFWRcan be insufficient

is its uneven spatial distribution. Annual runoff

(Fig. 2A) can be considered as the maximum

available RFWR if water from upstream cannot

be reused downstream because of consumptive

F R E S H W A T E R R E S O U R C E S

1Institute of Industrial Science, The University of Tokyo,4-6-1 Komaba, Meguro-ku, Tokyo 153-8505, Japan. 2JapanScience and Technology Agency, 4-1-8 Honcho, Kawaguchi-shi, Saitama 332-0012, Japan. 3Frontier Research Centerfor Global Change, Japan Agency for Marine-Earth Scienceand Technology, 3173-25 Showa, Kanazawa-ku, Yokohama236-0001, Japan. 4Research Institute for Humanity andNature, 457-7 Motoyama, Kamigamo, Kita-ku, Kyoto 603-8047, Japan.

*These authors contributed equally to this work.†To whom correspondence should be addressed. E-mail:taikan@iis.u-tokyo.ac.jp

25 AUGUST 2006 VOL 313 SCIENCE www.sciencemag.org1068

on

June

7, 2

013

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

use or water pollution (11). Runoff is accumu-

lated through river channels and forms river

discharge (Fig. 2B). River discharge can be con-

sidered as the potentially maximum available

RFWR if all the water from upstream can be

used. Both runoff and river discharge are con-

centrated in limited areas, and the amounts range

from nearly zero in desert areas through more

than 2000 mm/year of runoff in the tropics and

more than 200,000 m3/s of discharge on average

near the river mouth of the Amazon. Further-

more, the water demands for ecosystems and

navigation should also be met, and all the

RFWR cannot be used only for human beings.

How Are the World Water Resources Assessed?

In the late 1960s, the International Hydrological

Decade promoted studies on world water bal-

ances, and pioneering estimates were published

in the 1970s (5, 12, 13). Shiklomanov (4) as-

sembled country statistics on water withdrawals

in the past and present and made future

projections. Recent advances in information tech-

nologies have enabled global water-balance esti-

mations at finer spatial resolution (11, 14, 15).

Water withdrawals now can be distributed into

grid boxes, using the distributions of popula-

tion and the irrigation area as proxies, and

compared with the available RFWR in each

grid box (11, 14, 15).

The water scarcity index is defined as Rws 0(W – S)/Q, where W, S, and Q are the annual

water withdrawal by all the sectors, the water

use from desalinated water, and the annual

RFWR, respectively. A region is usually con-

sidered highly water stressed if Rws is higher

than 0.4 (7, 11, 14, 15). It is considered to be a

reasonable, although not definitive, threshold

value because not all the RFWR can be used

by human society. Data with shorter time scales

will enable more detailed assessments consider-

ing the effects of temporal variability in the

hydrological cycles.

In the era of the ‘‘Anthropocene’’ (16), where

human impacts on natural processes are large

and widespread, it no longer makes sense to

study only natural hydrological cycles. For this

reason, some studies have started to consider the

impact of human interventions on the hydrolog-

ical cycles, thereby simulating more realistically

the hydrological cycles on a global scale. In such

studies, human withdrawals are subtracted from

the river flow (15), and the regulation of flow

regime by major reservoirs is incorporated (17).

The distribution of the water scarcity index

Rws (11), recalculated with the latest multimodel

ensemble estimates (3), is shown in Fig. 2C. Rws

is high in Northern China, in the area on the

border between India and Pakistan, in the Middle

East, and in the middle and western areas of the

United States. Based on this assessment, approx-

imately 2.4 billion people are currently living in

highly water-stressed areas (18).

Can the ‘‘Virtual Water Trade’’ Alone Save theWater-Stressed Regions?

Transporting water over long distances, from

regions where water is abundant to dry regions

under water stress, is only feasible when gravity

can be used. The demand for high-quality drink-

ing water is limited to a few liters per person per

day and can be met through international trade

or by desalination. However, other demands for

water for households, industry, and agriculture

require up to one metric ton of water per day per

Fig. 1. Global hydrological fluxes (1000 km3/year) and storages (1000 km3)with natural and anthropogenic cycles are synthesized from various sources(1, 3–5). Big vertical arrows show total annual precipitation and evapo-transpiration over land and ocean (1000 km3/year), which include annual

precipitation and evapotranspiration in major landscapes (1000 km3/year)presented by small vertical arrows; parentheses indicate area (million km2).The direct groundwater discharge, which is estimated to be about 10% oftotal river discharge globally (6), is included in river discharge.

SPECIALSECTION

www.sciencemag.org SCIENCE VOL 313 25 AUGUST 2006 1069

on

June

7, 2

013

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

person in developing countries and considera-

bly more in developed countries. Therefore, the

supply for these sectors must be inexpensive,

which means that transporting water by tanker

or other high energy–consuming means is

generally not realistic (8).

On the other hand, water demand for food

and industrial production in dry regions can be

offset by importing food or industrial goods. Such

trade is called ‘‘virtual water trade’’ (19–21). The

weight of traded goods is normally just a

small fraction, such as 1/100 to 1/1000, of the

weight of the water required to produce that

goods, so transporting goods is considerably

easier than transporting the water itself. Total

international ‘‘virtual water trade’’ is estimated

to be about 1000 km3/year (20, 21), although

only a part of that ‘‘virtual water trade’’ is

done to compensate for water shortage (22).

Problems of water, food, health, and poverty

are interlinked in many developing countries,

particularly in the regions where freshwater

resources are scarce, the local economy is too

weak to allow import of food from outside on a

large scale, and desalination plants are im-

practical to implement.

On the contrary, once water supply is se-

cured by appropriate infrastructure investments

and improved management, public health con-

ditions improve, food supply stabilizes, the po-

tential for industrial development increases, and

the time that was earlier devoted to acquiring

water can be used for more productive work or

educational opportunities. This is the reason

that the target ‘‘Reduce by half the proportion of

people without sustainable access to safe

drinking water’’ (23) is one of the Millennium

Development Goals of the United Nations.

How Will Water Use Change in the Future?

The global population will certainly grow, at

least for several decades, and water demand

will increase as a result. Water demand per

person will most likely also increase due to

economic growth. For example, an expected

growth of meat consumption will increase the

water demand for fodder production.

The ultimate objectives of future-oriented

world water resource assessments are to show

the international community what will happen

if we continue to manage our water resources as

we do today and to indicate what actions may

be needed to prevent undesirable outcomes. In

that sense, studies of future world water re-

sources are successful if their predictions based

on business-as-usual are proven wrong. In line

with this, plausible scenarios informed by past

experiences and current trends are built for

future projections of the demand side.

In the agricultural sector, which is estimated

to withdraw two-thirds of world water with-

drawals and which accounts for 90% of total

water consumption in the world (4), in the

period from 1961 to 2004, crop yield per area

increased by a factor of 2.3, more than the rate of

population growth (2.0), and the total crop yield

increased by a factor of 2.4, even though the

area of cropland increased by only 10% and

harvested area increased less than that (24).

This phenomenal growth was to a large extent

due to a doubling of the irrigation area and the

Fig. 2. Global distribution of (A) mean annual runoff (mm/year), (B) mean annual discharge(million m3/year), and (C) water scarcity index Rws (3, 11). Water stress is higher for regions withlarger Rws.

F R E S H W A T E R R E S O U R C E S

25 AUGUST 2006 VOL 313 SCIENCE www.sciencemag.org1070

on

June

7, 2

013

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

corresponding increase of water withdrawal for

irrigation in addition to the increased usage of

fertilizer. Domestic per capita water use has

increased with gross domestic product (GDP)

growth, but in many developed countries this

increase seems to have come to an end; in some

countries, domestic per capita water use is now

decreasing. Such trend shifts should be con-

sidered in predictions of future water use. In-

dustrial water use has also increased along with

GDP, but recycling technology has reduced the

net intake of water for factories. For example,

nearly 80% of water used in the industrial sec-

tors in Japan is currently recycled (25).

There are concerns that in the decades

ahead, water withdrawals for irrigation cannot

be increased as required and that the lack of

water will impede necessary growth of food

production. However, Rws in developing coun-

tries is generally low, which means that they

should have a potential to increase their water

withdrawals. A key challenge for these coun-

tries should be how to implement soft mea-

sures (such as legislation, policies, and market

mechanisms) in addition to technical ones to

simultaneously increase the supply and man-

age the demand wisely (26).

What Effects Will Climate ChangeHave on RFWR?

The effect of global climate change on hy-

drological cycles is still uncertain, but higher

temperatures will turn some part of snowfall

into rainfall, the snowmelt season will be

earlier, and, as a result, the timing and volume

of spring flood will change substantially (27).

Nearly half of the world’s population depends

on groundwater sources for drinking water

supply and for other uses (28). Sea level rise

will cause saline water intrusion into ground-

water aquifers near the coasts and will decrease

the available groundwater resources. On the

water demand side, changes in the seasonal

pattern have not been estimated globally, and a

comprehensive description of groundwater

withdrawal in the world is largely lacking.

Lack of seasonal details in existing assess-

ments is the reason that crude annual average

measures such as the water scarcity index Rws

and the Falkenmark’s indicator or the ‘‘water

crowding indicator’’ Aw 0 Q/C (4), where C is

population size, are still being used. Of course,

there have been advances in world water re-

source assessments; projections on the demand

side now are based on the Intergovernmental

Panel on Climate Change’s Special Report on

Emissions Scenarios, making them consistent

with future climate projections (18, 29, 30),

and uncertainties in the projections of future

hydrological cycles are reduced by the use of

multimodel ensemble technique (18, 31, 32).

Figure 3 compares three assessments of the

number of people who will live in regions with

high water stress until the end of the 21st

century (11, 18, 29, 30). Even though the

projections vary by scenario, their estimates

correspond fairly well. Notably, climate

change is expected to accelerate the global

hydrological cycles, and precipitation will

increase on average. Evapotranspiration will

not increase as much as precipitation globally

because elevated CO2concentration induces

stomata closure and reduces transpiration

(33), and river discharge will increase on

global scale because of the increased precip-

itation and the reduced transpiration (31, 32).

As a result, the available RFWR is expected

to increase at a higher rate than water de-

mand, calculated from population and eco-

nomic growth. Because of this, the water

scarcity index Rws and the water crowding

indicator Aw, both based on annual RFWR,

show that water stress will be reduced on a

macro scale, except for the A2 scenario,

which represents a very heterogeneous world

with high population growth. However, the

decrease in the number of people under water

stress is only marginal and the results should

not be viewed too optimistically because they

are based on estimates of annual RFWR.

Other anticipated impacts of climate change

on water resources, such as modification of

seasonal variation of available RFWR, deg-

radation of water quality, and associated

changes in water resource management, are

not taken into account. Furthermore, precip-

itation will become more intense and inter-

mittent, and the risks of floods and droughts

will increase, sometimes in the same region

of the world (34). These changes in risks are

not well considered in current global assess-

ments on future water resource management.

Nevertheless, it is certain that there are people

who are already suffering from water shortage

today and that any change in the hydrological

cycle will demand changes in water resource

management, whether the change is caused by

global warming or cooling, or by anthropogenic

or natural factors. If society is not well prepared

for such changes and fails to monitor variations

in the hydrological cycle, large numbers of

people run the risk of living under water stress

or seeing their livelihoods devastated by water-

related hazards such as floods.

How Can Hydrological Science HelpSolve World Water Issues?

Detailed knowledge of global water resources

certainly has been enriched over the 40 years

that have passed since the International Hydro-

logical Decade. Water cycles on Earth can now

be measured and simulated on finer temporal

and spatial scales with detailed models of each

hydrological process, and the current and future

status of the global water system can be il-

lustrated (Figs. 1 to 3). In contrast to these

achievements in studies of the natural hydro-

logical cycles, data about the social aspects of

water use are not easily available.

Finally, the future development of hydrolo-

gy requires improved communication between

scientists and policy-makers to ensure that

hydrological expertise is translated into actions

Fig. 3. Current and future projections of population under high water stress under three business-as-usual scenarios of the Intergovernmental Panel on Climate Change’s Special Report on EmissionsScenarios. Threshold values are set to be (A) the water-crowding indicator Aw 0 Q/C G 1000 m3/year percapita and (B) the water scarcity index Rws 0 (W – S)/Q 9 0.4, where Q, C, W, and S are renewablefreshwater resources (RFWR), population, water withdrawal, and water generated by desalination,respectively. Error bars indicate the maximum and minimum population under high water stresscorresponding to the RFWR projected by six climate models. Climatic conditions averaged for 30 yearsare used for the plots at 2025 (averaged for 2010 to 2039), 2055 (averaged for 2040 to 2069), and2075 (averaged for 2060 to 2089).

SPECIALSECTION

www.sciencemag.org SCIENCE VOL 313 25 AUGUST 2006 1071

on

June

7, 2

013

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

that address water challenges (35) and to make

sure that scientists understand what kinds of

knowledge are required by policy-makers and

by society at large.

References and Notes1. T. Oki, in Encyclopedia of Hydrological Sciences,

M.G. Anderson, J. McDonnell, Eds. (Wiley, New York,2005), vol. 1, pp. 13–22.

2. M. W. Rosegrant, C. Ximing, S. A. Cline, Environment 45,24 (2003).

3. P. A. Dirmeyer et al., Bull. Am. Meteorol. Soc., in press.4. I. A. Shiklomanov, Ed., Assessment of Water Resources

and Water Availability in the World (World MeteorologicalOrganization/Stockholm Environment Institute, Geneva,Switzerland, 1997).

5. V. I. Korzun, World Water Balance and Water Resources ofthe Earth, vol. 25 of Studies and Reports in Hydrology(UNESCO, Paris, 1978).

6. T. M. Church, Nature 380, 579 (1996).7. M. Falkenmark, J. Rockstrom, Balancing Water for

Humans and Nature (Earthscan, London, 2004).8. S. L. Postel, G. C. Daily, P. R. Ehrlich, Science 271, 785

(1996).9. C. Nilsson, C. A. Reidy, M. Dynesius, C. Revenga, Science

308, 405 (2005).10. World Register of Dams (International Commission on

Large Dams, Paris, 1998).

11. T. Oki et al., Hydrol. Sci. J. 46, 983 (2001).12. M. I. Lvovitch, Trans. Am. Geophys. Union 54, 28

(1973).13. F. Baumgartner, E. Reichel, The World Water Balance:

Mean Annual Global, Continental and MaritimePrecipitation, Evaporation and Runoff (Ordenbourg,Munchen, Germany, 1975).

14. C. J. Vorosmarty, P. Green, J. Salisbury, R. B. Lammers,Science 289, 284 (2000).

15. J. Alcamo et al., Hydrol. Sci. J. 48, 339 (2003).

16. P. J. Crutzen, Nature 415, 23 (2002).17. N. Hanasaki, S. Kanae, T. Oki, J. Hydrol. 327, 22

(2006).18. N. Utsumi, thesis, The University of Tokyo (2006).19. J. A. Allan, Yale Univ. For. Environ. Studies Bull. 103, 141

(1998).

20. T. Oki et al., in Value of Water Research Report Series No.12, A. Y. Hoekstra, Ed. (UNESCO-IHE, Delft, Netherlands,2003), pp. 221–235.

21. A. K. Chapagain, A. Y. Hoekstra, Value of Water ResearchReport Series No. 13 (UNESCO-IHE, Delft, Netherlands,2003).

22. H. Yang, P. Eichert, K. C. Abbaspour, A. B. Zehnder,Environ. Sci. Technol. 37, 3048 (2003).

23. UN Millennium Development Goals (www.un.org/millenniumgoals/index.html).

24. FAOSTAT (http://faostat.fao.org/).25. Water Resources in Japan (Ministry of Land, Infrastructure,

and Transport, Tokyo, 2005).

26. P. Gleick, Science 302, 1524 (2003).27. Intergovermental Panel on Climate Change, Climate

Change, 2001: The Science Basis (Cambridge Univ. Press,Cambridge, UK, 2001).

28. World Water Day (www.worldwaterday.org/wwday/1998/).29. J. Alcamo et al., personal communication.30. N. W. Arnell, Glob. Environ. Change 14, 31 (2004).31. D. Nohara, A. Kitoh, M. Hosaka, T. Oki, J. Hydromet.,

in press.32. P. C. D. Milly, K. A. Dunne, A. V. Vecchia, Nature 438,

347 (2005).33. N. Gedney et al., Nature 439, 835 (2006).34. Y. Hirabayashi et al., in preparation.35. T. Oki, C. Valeo, K. Heal, Eds., Hydrology 2020: An

Integrating Science to Meet World Water Challenges,IAHS Publication 300 (IAHS, Wallingford, UK, 2005).

36. We thank our research group members at Institute ofIndustrial Science, The University of Tokyo for comments andsupport, particularly Y. Shen, M. Bengtsson, Y. Agata, H. Kim,and N. Utsumi. We also thank the anonymous reviewers forcomments. For financial support, we thank Core Research forEvolutional Science and Technology (CREST), the JapanScience and Technology Agency (JST), the Research Institutefor Humanity and Nature (RIHN), the special coordinationfunds for promoting science and technology by the JapanCabinet Office, and Global Environment Research Fund(GERF) of the Ministry of the Environment of Japan.

10.1126/science.1128845

REVIEW

The Challenge of Micropollutantsin Aquatic SystemsRene P. Schwarzenbach,* Beate I. Escher, Kathrin Fenner, Thomas B. Hofstetter,C. Annette Johnson, Urs von Gunten, Bernhard Wehrli

The increasing worldwide contamination of freshwater systems with thousands of industrial andnatural chemical compounds is one of the key environmental problems facing humanity. Althoughmost of these compounds are present at low concentrations, many of them raise considerabletoxicological concerns, particularly when present as components of complex mixtures. Here wereview three scientific challenges in addressing water-quality problems caused by such micro-pollutants. First, tools to assess the impact of these pollutants on aquatic life and human healthmust be further developed and refined. Second, cost-effective and appropriate remediation andwater-treatment technologies must be explored and implemented. Third, usage and disposalstrategies, coupled with the search for environmentally more benign products and processes,should aim to minimize introduction of critical pollutants into the aquatic environment.

About one-fifth of the world_s popula-tion does not have access to safe water,

and two-fifths suffer the consequences

of unacceptable sanitary conditions (1). Patho-

gens in water cause more than 2 million deaths

annually; most are children under the age of 5.

The increasing chemical pollution of surface

and groundwaters, with largely unknown long-

term effects on aquatic life and on human

health, could easily lead to a problem of similar

or even greater magnitude. More than one-third

of the Earth_s accessible renewable freshwateris used for agricultural, industrial, and domes-

tic purposes, and most of these activities lead

to water contamination with numerous syn-

thetic and geogenic compounds (Table 1). It

therefore comes as no surprise that chemical

pollution of natural waters has already become

a major public concern in almost all parts of the

world.

Industry and municipalities use about 10%

of the globally accessible runoff and generate a

stream of wastewater, which flows or seeps into

rivers, lakes, groundwater, or the coastal seas

(1). These wastewaters contain numerous chem-

ical compounds in varying concentrations. About

300 million tons of synthetic compounds annu-

ally used in industrial and consumer products

partially find their way into natural waters (Table

1). Additional pollution comes from diffuse

sources from agriculture, where 140 million

tons of fertilizers and several million tons of

pesticides are applied each year (2). In the

European Union, for instance, there are more

than 100,000 registered chemicals, of which

30,000 to 70,000 are in daily use (EINECS,

European Inventory of Existing Chemical Sub-

stances). The input of 0.4 million tons of oil and

gasoline components through accidental spills

represents yet another important source of water

pollution. Other notable sources of contamination

are the intrusion of salty water into groundwater

due to overexploitation of aquifers; the human-

driven mobilization of naturally occurring geo-

genic toxic chemicals, including heavymetals and

metalloids (Table 1); and the biological produc-

tion of toxins and malodorous compounds.

To date, an effective and sustainable global

strategy against this insidious and mostly unseen

contamination of aquatic environments barely ex-

ists. Source controls and technical systems, such

as wastewater treatment plants, function as par-

tial barriers, particularly in highly industrialized

countries, but major challenges remain. The

source, behavior, and treatment of the relatively

small number of macropollutants (3) such as

acids, salts, nutrients, and natural organic matter,

occurring at mg/liter to mg/liter concentrations,

are relatively well understood: High nutrient

loads can lead to increased primary production,

Eawag, Swiss Federal Institute of Aquatic Science andTechnology, 8600 Dubendorf, Switzerland, and Institute ofBiogeochemistry and Pollutant Dynamics, ETH Zurich,8092 Zurich, Switzerland.

*To whom correspondence should be addressed. E-mail:rene.schwarzenbach@env.ethz.ch

F R E S H W A T E R R E S O U R C E S

25 AUGUST 2006 VOL 313 SCIENCE www.sciencemag.org1072

on

June

7, 2

013

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

www.sciencemag.org SCIENCE VOL 338 7 DECEMBER 2012 1273

NEWS&ANALYSISC

RE

DIT

S (T

OP

TO

BO

TT

OM

): R

ICH

AR

D K

ING

SF

OR

D; D

ISC

OV

ER

MU

RR

AY

RIV

ER

, W

WW

.MU

RR

AY

RIV

ER

.CO

M.A

U

MELBOURNE, AUSTRALIA—In the scramble

to secure water for drinking and agricul-

ture, the environment often gets overlooked.

Australia is aiming to set an example of how

to balance all three needs in its agricultural

heartland. The Murray-Darling basin plan,

adopted by the government on 22 November,

“is historic,” says Richard Kingsford, a con-

servation biologist at the University of New

South Wales in Sydney. South Australian

River Murray Minister Paul Caica praised the

expected environmental gains: “It will help

fl oodplains support healthy red gum forests,

waterbird and fi sh breeding habitats. … It

will keep water levels high enough to prevent

acidifi cation in the lower lakes …

and reduce the risk of the Murray

[m]outh needing to be dredged.”

Water managers worldwide

“see it as a model” for the use

of technical advice and involve-

ment of government at all levels,

says water policy expert Sharon

Megdal of the University of Ari-

zona in Tucson. The plan’s deft

balancing of societal and ecolog-

ical needs, adds Clifford Dahm,

an aquatic ecologist at the Uni-

versity of New Mexico, Albu-

querque, “gives us some useful

ideas for science, planning, and

the political tightrope that such

decisionmaking requires.”

The 1 -mi l l ion - square -

kilometer Murray-Darling basin

stretches across four states that

have been brawling over water

for more than a century. Rising

from tributaries in Queensland

and New South Wales, the Dar-

ling River flows southwest,

joining the Murray River at the

Victorian border, and f inally

enters the sea near Adelaide.

Although the two rivers and their

tributaries account for less than

7% of the nation’s stream fl ow,

the basin, which draws most of

its water from the Murray, pro-

duces about 40% of the food

grown in Australia. Its ecological

treasures include Chowilla wet-

lands, Macquarie marshes, and,

near the mouth of the Murray, the

iconic Coorong with its 140-kilometer-long

lagoon that provides sanctuary for a cornu-

copia of waterbirds.

Beginning in the 1950s, dams and irriga-

tion channels transformed the free-fl owing

waterways into a highly managed system.

Prior to the arrival of Europeans in the 1800s,

more than 40% of rainwater entering the sys-

tem annually, or 12,200 billion liters, made

it to the sea, according to a 2008 study by

the Commonwealth Scientific and Indus-

trial Research Organisation. By 2008, even

with normal rainfall, fl ow in the rivers’ lower

reaches had dwindled to nearly one-third of

historical levels, about 4700 billion liters. The

decadelong Millennium Drought starting in

the late 1990s exacerbated the water short-

age. Along the lower Murray, three-quarters

of once-extensive red gum forests withered

and died. The Goolwa wetlands dried up,

oxidizing sulfates in the soil to sulfuric acid,

which acidified the wetlands and adjacent

farms once the waters returned and spread

it. The Coorong was cut off from the Mur-

ray for 3 years. Its salinity jumped to fi ve to

seven times that of seawater, turning half of

the estuary into a dead sea. “Paleoecologi-

cal evidence indicates it was unusual even

for the last 7000 years,” Rebecca Lester, an

environmental scientist at Deakin University

in Melbourne, penned on The Conversation,

a research news Web site.

Upstream, nutrients accumulating in the

Murray and Darling rivers fueled toxic algal

blooms, setting a world record in 1991 for a

bloom that extended 1000 kilometers along

the Darling River. Residents of Adelaide,

the city last in line to draw water

from the river system, had to

contend with some of the world’s

saltiest drinking water. In 2009,

experts raised the alarm with

some predicting that within the

year, the city would have to turn

to bottled water.

The Millennium Drought led

to a number of measures, includ-

ing the establishment in 2008 of

the Murray-Darling Basin Author-

ity (MDBA), which is an indepen-

dent federal agency charged with

fi nding the right balance between

agricultural, social, and environ-

mental values. Restoring river and

wetlands health meant primar-

ily removing less water for irri-

gation. “How much to take back

is open for lots of debate,” says

Tony Minns, director of the Goy-

der Institute for Water Research

in Adelaide. MDBA planners

also called for fi xing leaky irriga-

tion systems to retain more water

in rivers and proposed that fl ows

could be bolstered by pumping in

groundwater.

In June 2010, the Wentworth

Group of Concerned Scientists,

an independent conservation

organization, called for water

fl ows to be returned to two-thirds

of historical levels, a target that

won broad though not universal

support among environmental-

ists. MDBA issued a preliminary

plan in November 2010 that

River Basin Management Plan Secures Water for the Environment

AU ST R A L I A

DAR

LIN

G RIV

ER

Free-fl owing. Balancing

competing demands will keep

water fl owing through the

Murray-Darling river system.

Published by AAAS

on

June

7, 2

013

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om

7 DECEMBER 2012 VOL 338 SCIENCE www.sciencemag.org 1274

NEWS&ANALYSIS

CR

ED

IT: SA

RA

PR

ESS

looked at three water recovery scenarios—none of which pleased farmers, who tossed it into bonfi res. A fi nal plan issued a year later improved modeling to assess how dif-ferent scenarios would affect 18 key indica-tors of environmental health. One indicator is to keep salinity in the Coorong below 60 grams per liter—a safe level for seagrasses that fi sh and birds depend on. (Pre-European salinity was 24 grams per liter; by 2009 it had risen to 62 grams per liter.) Returning 3200 billion liters to the Murray-Darling system would satisfy 17 of 18 indicators and come close to the Wentworth target of two-thirds of historical fl ows. But it poses another problem: Infrastructure along the rivers’ lower reaches built to suit reduced flows would be swamped. To accommodate more fl ow, the government earmarked $1.85 bil-lion in October to raise bridges, build levees, and compensate landowners. Much of that

water will come from buying water rights from farmers; better irrigation infrastruc-ture will spread it further. The basin plan is expected to be fully implemented by 2024.

The plan still faces hurdles. “There are many ways the states could frustrate the deal,” says Rhondda Dickson, MDBA chief executive. For instance, states still own the water and could cap the amount of irriga-tion rights MDBA can purchase, as New South Wales has threatened to do. “To make it work,” Dickson says, “the government needs the states’ support, however grudging.” The environment ministry is working on an implementation agreement with the states.

Ecological recovery could take decades. MDBA’s water purchases and infrastructure upgrades so far have returned 1327 billion liters to the rivers, rejuvenating parched wet-land systems, Kingsford says. The numbers of nesting waterfowl species have bounced

back, but Ruppia tuberosa, a key seagrass species that fi sh and waterfowl feed on, has disappeared. Soil acidifi cation still affects wetlands and farm fields. “It’s not clear whether these changes are permanent,” Minns says, referring to both biodiversity loss and acidifi cation.

The basin’s future will depend not just on the volume of water fl owing through the sys-tem, but how it’s added. “We have to develop ecological watering plans [to decide] how much, how often and when [to release water] to achieve the environmental outcomes,” Minns says. One goal is to restore natural flow variability. “Ecologically, occasional fl ooding is important for these river systems,” Kingsford says. Now that MDBA is getting the water it needs, it will have to learn how to best manage it. As Minns says, “The real work starts now.”

–ELIZABETH FINKEL AND DENNIS NORMILE

Be bolder.A new report by a presidential advisory

panel urges U.S. research agencies to make a bigger commitment to supporting high-risk, interdisciplinary research by investiga-tors with a strong track record. Current efforts by the National Institutes of Health (NIH) and the National Science Foundation (NSF) to do so are “a drop in the bucket of total agency funding,” says a report released last week by the President’s Council of Advisors on Science and Technology (PCAST).

The new PCAST report, entitled Transformation

and Opportunity: The

Future of the U.S. Research

Enterprise, offers 17 rec-ommendations to shore up the $450-billion-a-year U.S. research enterprise, both public and private, and increase its eventual eco-nomic payoff. Most of the ideas—including the need to ease the regulatory burden on universi-ties, strengthen ties with industry, improve science and math education, and ensure a steady, predictable rate of growth for federal spending on science—will sound familiar to readers of other recent reports by equally prestigious panels. “The majority of these

recommendations have been made by other groups,” admits University of Texas, Aus-tin, computer scientist William Press, who co-chaired the working group that produced the report. (Press is also president of AAAS, which publishes Science.) In fact, PCAST’s suggestions to boost overall research spend-

ing to 3% of the nation’s gross domestic product, make permanent a research tax credit for industry, and help foreign scientists remain in the United States after graduation have already been embraced by its intended audi-ence: President Barack Obama. But advocates for the research community say that it never hurts to remind the administration that such proposed policy changes are important—especially when the White House and Congress are engaged in tense negotia-tions to avoid heading over

the so-called fi scal cliff next month.One area where the PCAST report breaks

from the pack is its call for research agencies to adopt “revolutionary … interdisciplinary ... and people-based awards.” Too much of federal spending—which amounts to about one-third of the nation’s total investment

in research—backs incremental advances, according to the report. And one of the cul-prits is a conservative merit review process that rewards safe bets.

NIH is taking small steps away from that approach with its Pioneer, New Innovator, and Early Independence awards from the director’s offi ce, the report notes. Likewise, the report praises NSF’s Rapid Response Research grants and its planned expansion of programs to reward “creative … transforma-tive interdisciplinary ventures.” But Press says “the funding agencies have been slower than we would like to see in moving in these direc-tions.” And the report documents those baby steps: This year, it notes, NIH will make 50 awards in the three director’s categories out of a total of 35,944 research grants. “While this plethora of initiatives, each worthy in its own way, gives an illusion of signifi cant progress,” the report notes, “in truth the sum of all these programs is tiny, almost invisible, in com-parison to each agency’s dominant” form of research support.

The report also urges Congress and the executive branch to find a way to provide research agencies with multiyear budgets, or at least funding guarantees for individual projects. Other countries operate on 5-year budgets, notes NSF Director Subra Suresh. But Suresh says “it’s very diffi cult to carry out long-term planning with our sister agen-cies” under the current U.S. system of annual appropriations. –JEFFREY MERVIS

White House Panel Urges Agencies to Take More Risks

U. S . S C I E N C E P O L I C Y

Presidential prodding. William

Press led the PCAST panel that wants

changes in U.S. research practices.

Published by AAAS

on

June

7, 2

013

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om